The long-term efficacy of pelvic floor exercise (PFE) as a treatment for urinary incontinence is threatened by patient compliance. To enhance compliance, PFE needs to be effective, while reducing the time burden placed on the patient. Our long-term objectives are to 1) validate a physiologically based, task-specific approach to PFE and 2) determine the exercise frequency needed to preserve muscle strength. Our specific aims are to determine: 1) the effectiveness of an individualized, physiologically based PFE regimen, 2) if a treatment progression to include exercise in upright postures is more effective than one performed in a lying position, and 3) if a low frequency exercise program is sufficient to preserve therapeutic gains. Over a six-month period, we will study 50 women, ages 40-70, with genuine stress incontinence. Based on symptom severity, we will randomly assign subjects to perform exercises either in upright postures or in a lying position. Subjects will attend 12, weekly physical therapy (PT) visits and perform a home exercise program (HEP) consisting of muscle strength, power, and endurance exercises. Using electromyography (an electrical recording of muscle activity), we will determine exercise progression using a protocol that individualizes exercise prescription based on muscle fatigue. At each weekly assessment, we will adjust the subject's HEP accordingly. Subjects assigned to the upright training group will perform one third of their exercises in the supine, sitting and standing positions. The other group will perform all exercises while in a lying position. At week 12, we will randomly assign subjects in each treatment group to either a low- (exercises once a week) or high (exercises every other day) maintenance group. To determine treatment outcomes, we will compare group differences in the number of incontinent episodes, amount of urine loss during a pad test, a measure of urethral resistance (Valsalva leak point pressure), and scores on a quality of life measure. To determine the overall effectiveness of the treatment protocol, we will compare pre- to post-treatment changes in incontinent episodes to similar data reported in the literature. We will also examine the relationships between menopausal status, incontinence severity, and treatment outcomes. The results of this study may lead to a more effective exercise approach. Thus, future clients may benefit from fewer PT visits, greater pelvic muscle function, and avoidance of surgery. A major factor threatening compliance is the burden of exercise. The investigators suggest that if they can demonstrate that reducing exercise frequency (thus reducing the burden of exercise) does not reduce the benefits of a formalized rehabilitation program, then long-term pelvic floor exercises compliance could be improved.